lable at ScienceDirect

Manual Therapy 17 (2012) 27e33

Contents lists avai

Manual Therapy

journal homepage: www.elsevier .com/math

Original article

The relationships between measures of stature recovery, muscle activityand psychological factors in patients with chronic low back pain

Sandra Lewis a,*, Paul Holmes a, Steve Woby b, Jackie Hindle c, Neil Fowler a

a Institute for Performance Research, Manchester Metropolitan University, Crewe CW1 5DU, United KingdombResearch & Development, Pennine Acute Hospitals NHS Trust, North Manchester General Hospital, Manchester M8 5RB, United KingdomcResearch Institute for Health and Social Change, Manchester Metropolitan University, Manchester M13 0JA, United Kingdom

a r t i c l e i n f o

Article history:Received 7 January 2011Received in revised form29 July 2011Accepted 2 August 2011

Keywords:Low back painPsychologyElectromyography

* Corresponding author. Tel.: þ44 161 247 5765; faE-mail address: s.lewis@mmu.ac.uk (S. Lewis).

1356-689X/$ e see front matter � 2011 Elsevier Ltd.doi:10.1016/j.math.2011.08.001

a b s t r a c t

Individuals with low back pain (LBP) often exhibit elevated paraspinal muscle activity compared toasymptomatic controls during static postures such as standing. This hyperactivity has been associatedwith a delayed rate of stature recovery in individuals with mild LBP. This study aimed to explore thisassociation further in a more clinically relevant population of NHS patients with LBP and to investigate ifrelationships exist with a number of psychological factors. Forty seven patients were recruited fromwaiting lists for physiotherapist-led rehabilitation programmes. Paraspinal muscle activity whilestanding was assessed via surface electromyogram (EMG) and stature recovery over a 40-min unloadingperiod was measured on a precision stadiometer. Self-report of pain, disability, anxiety, depression, pain-related anxiety, fear of movement, self-efficacy and catastrophising were recorded.

Correlations were found between muscle activity and both pain (r¼ 0.48) and disability (r¼ 0.43).Muscle activity was also correlated with self-efficacy (r¼�0.45), depression (r¼ 0.33), anxiety (r¼ 0.31),pain-related anxiety (r¼ 0.29) and catastrophising (r¼ 0.29) and was a mediator between self-efficacyand pain. Pain was a mediator in the relationship between muscle activity and disability. Staturerecovery was not found to be related to pain, disability, muscle activity or any of the psychologicalfactors. The findings confirm the importance of muscle activity within LBP, in particular as a pathway bywhich psychological factors may impact on clinical outcome. The mediating role of muscle activitybetween psychological factors and pain suggests that interventions that are able to reduce muscletension may be of particular benefit to patients demonstrating such characteristics, which may help inthe targeting of treatment for LBP.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Patients with low back pain (LBP) often demonstrate alteredmuscle function compared to asymptomatic controls. In particular,individuals with LBP have been found to exhibit hyperactivity of thesuperficial paraspinal muscles during static postures such asstanding (e.g. Ambroz et al., 2000).

The height of intervertebral discs changes in response tocompressive forces (due to a combination of fluid flow and elasticdeformation) and this is reflected in changes in stature. Staturechange is therefore used as a proxy measure of the load on thespine and measurements have been shown to correlate with moredirect measurements of changes in lumbar spine length assessed

x: þ44 161 247 6375.

All rights reserved.

via Magnetic Resonance Imaging (MRI) (Lewis and Fowler, 2009). Ithas also been shown that both chronic low back pain (CLBP)patients and asymptomatic individuals are able to produce staturemeasurements with a good level of repeatability (Healey et al.,2005b). Healey et al. (2005a) found significantly reduced staturerecovery in individuals with mild LBP compared to controls, withstature recovery negatively correlated with paraspinal muscleactivity. The authors hypothesized that the elevated muscle activityobserved in the LBP group resulted in greater compressive loads onthe spine that, in turn, prevented the intervertebral discs fromregaining their initial height and consequently prolonged staturerecovery. Reduced stature recovery may increase the risk of futureback pain and increase loading on spinal structures such as thefacet joints (Adams et al., 2002). Significant negative correlationsbetween stature recovery and both pain and disability appear tosupport the clinical relevance of this relationship (Healey et al.,2005a).

Fig. 1. EMG electrode placement.

S. Lewis et al. / Manual Therapy 17 (2012) 27e3328

Psychological factors are known to play an important role in LBPand are sometimes viewed as ‘obstacles to recovery’ (e.g. Fosteret al., 2010). It has been suggested that one of the ways psycho-logical factors may affect the condition is via increased spinalloading resulting from altered paraspinal muscle activity. Further-more, LBP patients with high levels of pain-related fear generallyexhibit elevated paraspinal muscle activity compared to low fearfulpatients (Vlaeyen et al., 1999), especially when confronted withmovements which they believe to be harmful (Vlaeyen and Linton,2000). It is proposed that pain-related fear may perpetuate painand disability via this muscle guarding. Muscle activity maytherefore be a contributory factor in the link between psychologicalfactors and clinical outcome.

The study carried out by Healey et al. (2005a) involved indi-viduals who all self-managed their pain. The aim of this studywas to extend those findings by analysing the relationshipbetween stature recovery, muscle activity, pain and disability ina more clinically relevant population of NHS patients with LBP,including individuals with more severe back pain than previouslyexamined. In addition, this study sought to establish whethera range of self-report psychological factors are associated withmuscle activity or stature change. An asymptomatic controlgroup was included in the design to enable comparison betweenthe two groups. It was hypothesized that, consistent with thefindings of Healey et al. (2005a), the patients with LBP wouldhave higher muscle activity and reduced stature recoverycompared to the asymptomatic group and that stature recoverywould be negatively related to each of muscle activity, pain anddisability. Based on the limited previous research in the area, itwas expected that muscle activity would be correlated with thepsychological factors considered. It was expected that thesepsychological factors may impact on stature recovery via theirinfluence on muscle activity and hence also lead to observedcorrelations with stature change.

2. Methods

2.1. Participants

Data were collected from 47 patients with LBP (age,46.2�11.1 yr; height, 166.4�7.5 cm; body mass, 79.3�18.8 kg)who had been referred to a physiotherapist-led rehabilitationprogramme in North Manchester and 18 asymptomatic controls(age, 44.6�13.3 yr; height, 169.0�10.3 cm; body mass,70.8� 12.7 kg). The LBP group was mixed (18 men, 29 women),between the ages of 23 and 70 years.

Patients on the waiting list for the rehabilitation programmewere sent information regarding the study through the post andasked to return a reply slip if they wished to participate. Theexclusion criteria consideredwere: nerve root compression; centralnervous system impairment; progressive motor deficit; sphincterimpairment from neurologic cause; and presence of ‘red flags’ (e.g.unexplained weight loss, recent urinary tract infection, history ofintravenous drug use). Many of the patients were taking analgesicsfor their back pain; it was not considered practical to exclude thoseon medication. Participants were offered £12.50 for each sessionthey attended to cover travel and parking expenses. Ethicalapproval was granted by the North Manchester NHS ethicscommittee and local NHS permissionwas granted by Pennine AcuteNHS Trust. All participants provided written informed consent.

The control group consisted of 9 men and 9 women agedbetween 25 and 64 years. Participants were excluded from thecontrol group if they had, at any time, experienced recurring orpersistent back pain, lost aworking day because of back pain, or hadconsulted a physician about back pain within the last 15 years.

2.2. Muscle activity measurement

Raw electromyographic signals were recorded using a DELSYSEMG system (Delsys Inc. Boston, MA, USA). Single differentialsurface electrodes consisting of two silver bars with an inter-electrode spacing of 10 mm were used. Signals were band-passfiltered between 20 and 450 Hz with a sampling frequency of1000 Hz.

As shown in Fig. 1, electrodes were placed over the erectorspinae muscle at the level of the L1e2 and L4e5 interspaces,approximately 3 cm from the midline on either side and thereference electrode was placed on the right iliac crest. Participantsthen assumed a standing posture for 10 s while a recording wastaken. The electromyogram (EMG) datawere normalised relative toa reference voluntary contraction (RVC). This reference taskrequired each participant to stand while holding up a specific mass(0.5 kg) in each hand with arms bent (upper arms horizontal, lowerarms vertical) for 10 s (shown in Fig. 2). The signal mean value wasremoved from the raw EMGs, before rectifying and integrating overa period of 5 s. The EMG reading and the RVC were both taken to bethe average of the three readings recorded during the session.

2.3. The stadiometer

Changes in stature were measured with a standing stadiometer,which consisted of a rigid frame, mounted at a right angle to a baseplate and inclined backward 15� from the vertical (Fig. 3). Fouranatomical points were identified (Lewis and Fowler, 2009) andthen supported by the frame to maintain the natural contours ofthe head and spine. The position of the feet was marked and head

Fig. 2. Participant performing a reference voluntary contraction (RVC). Fig. 3. Participant in position in the stadiometer.

S. Lewis et al. / Manual Therapy 17 (2012) 27e33 29

position was controlled by the use of spectacle frames withattached lasers, which were aligned with two movable targetsabove the participant’s head. A high-resolution linear variabledisplacement transducer (LVDT) (Solartron Metrology, DC50) wasused to detect changes in stature by measuring vertical displace-ment with an accuracy of approximately 0.1 mm. The informationwas observed graphically on a laptop computer at the time ofcollection and stored digitally for later analysis at a sampling rate of100 Hz.

All participants initially undertook a brief familiarisation sessionon the stadiometer to enable them to practice the adoption ofa repeatable and comfortable posture. This consisted of fiverecordings, between which the participant was asked to leanforward and break contact with the postural controls beforeresuming their position for the next measurement. A pilot studywas carried out in which eight asymptomatic participants all per-formed this brief familiarisation of five readings. The results(average standard deviation (SD) 1.0 mm, standard error ofmeasurement (SEM) 0.8 mm) demonstrated that this approach wassufficient to produce reliable stadiometer readings. Participantsremained in position for a period of 20 s and the stature value usedwas the mean reading over the final 10 s.

2.4. Measures

At the end of the testing session patients were asked about theirpain intensity during the past 24 h. They were then given a ques-tionnaire booklet containing a series of self-report measures.Although some patients completed the booklet immediately, themajority completed it at home and returned it at a later date.Asymptomatic controls completed only the anxiety and depression(HADS) questionnaire.

2.5. Pain intensity

A numerical rating scale (NRS) was employed to assess painintensity. Participants were asked to rate their pain during the past24 h on a scale ranging from (0) ‘no pain’ to (10) ‘worst possiblepain’. Research supports the reliability and validity of NRSs of painintensity (Jensen, 2003) and the 11-point NRS was recently rec-ommended by the Initiative on Methods, Measurement and PainAssessment in Clinical Trials (IMMPACT) to assess chronic painintensity (Dworkin et al., 2005).

2.6. Disability

The Roland Disability Questionnaire (RDQ; Roland and Morris,1983) is a 24-item self-report measure that assesses disabilitydue to back pain. Patients are asked to select which statements,related to perceived limitations in typical daily activities, apply tothem. The score is calculated by adding up the number of selecteditems, so that a higher number represents greater perceiveddisability. The RDQ has excellent reliability, validity and respon-siveness (Roland and Fairbank, 2000; Turner et al., 2003) and iswidely used to sample within the specific population for this study.

2.7. Anxiety and depression

The Hospital Anxiety and Depression Scale (HADS; Zigmond andSnaith, 1983) is a widely used measure, designed for use withgeneral medical outpatient populations. It consists of anxiety anddepression subscales, both of which have seven items. Each item israted on a four-point scale from 0 to 3, determining the extent towhich the individual feels that each statement applies to them.Higher scores therefore represent higher levels of anxiety and

Table 1Group characteristics (n¼ 47 patients, 18 controls).

Variable Patient group Control group Effectsize

Possiblerange

Mean (�SD) Mean (�SD)

Body mass (kg) 79.3� 18.8 70.8� 12.7* 0.45Stature change (mm) 2.8 (�2.6) 3.9 (�1.5) 0.42Muscle activity

(% of RVC)75.5 (�12.2)a 66.8 (�15.8)* 0.71

Disability 11.5 (�5.9) 0e24Pain NRS 4.8 (�2.2) 0e10Anxiety 8.5 (�3.7) 5.2 (�2.8)** 0.88 0e21Depression 6.3 (�3.4) 3.5 (�2.1)** 0.80 0e21Functional self-efficacy 45.5 (�18.3) 0e72Pain-related anxiety 42.4 (�23.9) 0e100Catastrophising 21.1 (�12.8) 0e52Fear of movement 36.7 (�9.3) 17e68

SD: standard deviation; RVC: reference voluntary contraction; NRS: numericalrating scale.*p< 0.05, **p< 0.01.

a The muscle activity details for the patient group exclude two outliers. If thesetwo patients’ details are included, the mean (�SD) would be 73.1% (�16.3%),p¼ 0.09, effect size¼ 0.38.

S. Lewis et al. / Manual Therapy 17 (2012) 27e3330

depression. Both subscales of the HADS have established validityand reliability in a clinical population (Johnston et al., 2000;Bjelland et al., 2002; Pallant and Bailey, 2005).

2.8. Functional self-efficacy

Functional self-efficacy refers to the confidence that an indi-vidual has in their ability to successfully accomplish functionalactivities. The functional subscale of the Chronic Pain Self-EfficacyScale (CPSS-PF; Anderson et al., 1995) was used to assess func-tional self-efficacy. The original CPSS-PF, which has been shown tobe reliable and valid (Anderson et al., 1995), is scored on a ten-pointLikert scale, with higher scores reflecting higher levels of self-efficacy. This study used a nine-point Likert scale because itprovides patients with a mid-point option and has also been shownto have excellent internal consistency and testeretest reliabilitywith CLBP patients (Woby et al., 2007).

2.9. Fear of movement/(re)injury

The Tampa Scale of Kinesiophobia (TSK; Kori et al., 1990) aims tomeasure fear of movement/(re)injury in individuals with pain. Itconsists of 17 items, for which patients rate themselves on a four-point Likert scale ranging from ‘strongly disagree’ to ‘stronglyagree’. Four of the items are inversely scored. Scores range from 17to 68, with higher scores representing greater fear of movement/(re)injury. The English version of the TSK has been found to be validand reliable (Woby et al., 2005; French et al., 2007).

Table 2Correlation coefficients between outcome measures (n¼ 47).

1. 2. 3. 4.

1. Stature change e

2. Muscle activitya �0.15 e

3. Disability �0.16 0.43** e

4. Pain �0.16 0.48** 0.57** e

5. Anxiety �0.13 0.31* 0.55** 0.42**6. Depression 0.03 0.33* 0.57** 0.34*7. Self-efficacy 0.02 �0.45** �0.73** �0.49**8. Pain-related anxiety �0.08 0.29* 0.54** 0.37*9. Catastrophising 0.07 0.29* 0.41** 0.38**10. Fear of movement �0.14 0.20 0.39** 0.27*

*p< 0.05, **p< 0.01.a The muscle activity data for the patient group exclude two outliers.

2.10. Catastrophising

The Pain Catastrophising Scale (PCS; Sullivan et al., 1995)consists of 13 items describing thoughts or feelings that may beexperienced when in pain. For each statement, participants rate theextent to which they have the feeling when experiencing pain ona five-point Likert scale ranging from (0) ‘not at all’ to (4) ‘all thetime’. A higher score indicates greater levels of pain catastrophis-ing. The PCS has been shown to have good reliability and validity(Sullivan et al., 1995; Osman et al., 1997, 2000).

2.11. Pain-related anxiety

The Pain Anxiety Symptoms Scale-20 (PASS-20) is a shortened20-item version of the original Pain Anxiety Symptoms Scale (PASS:McCracken et al., 1992), a 40-item measure developed to assesspain-related anxiety in individuals with chronic pain. Higher scoresrepresent higher levels of pain-related anxiety. The PASS-20 hasbeen shown to have good factorial validity, internal consistency andclose association to the original 40-item version (Coons et al., 2004;Roelofs et al., 2004).

2.12. Procedure

Following the brief familiarisation session on the stadiometer,a baseline stature measurement and initial EMG readings (at restand during the RVC) were taken. Participants then assumed anunloading position on a physiotherapy bed for 20 min. This waseither a side-lying or a prone position, whichever the participantfelt was the most comfortable. After 20 min, the participants stoodup and performed the same EMG and stadiometer measurements,before again assuming an unloading position for a further 20 min.The same measurements were then taken for a final time. Staturechange was calculated as the difference between the final and theinitial stadiometer readings.

2.13. Analysis

Parametric tests were used, based on the results of Kolmogor-oveSmirnov and ShapiroeWilk tests of normality. One-tailed t-tests were calculated to examine for differences in levels of muscleactivity, stature recovery, anxiety and depression between the LBPpatients and the asymptomatic controls. For the LBP patients,Pearson’s correlation coefficient was calculated to investigate therelationships between the measures.

Mediational analysis (Baron and Kenny, 1986), including Sobel’stest of mediation, was carried out to investigate whether muscleactivity acted as a partial mediator between a range of psycholog-ical factors and either pain or disability, and also whether pain

5. 6. 7. 8. 9. 10.

e

0.68** e

�0.47** �0.57** e

0.67** 0.57** �0.56** e

0.59** 0.63** �0.45** 0.81** e

0.57** 0.44** �0.25 0.63** 0.65** e

R2 = 0.23p < 0.01

0

2

4

6

8

10

0% 20% 40% 60% 80% 100%

EMG (% of RVC)

Pai

n (N

RS)

Fig. 4. Relationship between pain NRS and muscle activity (as a % of the RVC). NRS:numerical rating scale; RVC: reference voluntary contraction.

R2 = 0.19p < 0.01

0

4

8

12

16

20

24

0% 20% 40% 60% 80% 100%

EMG (% of RVC)

Dis

abili

ty (RDQ)

Fig. 5. Relationship between disability (RDQ) and muscle activity (as a % of the RVC).RDQ: Roland Disability Questionnaire; RVC: reference voluntary contraction.

S. Lewis et al. / Manual Therapy 17 (2012) 27e33 31

mediated the effect between muscle activity and disability. It wasverified that there were no cases with a Cook’s distance of greaterthan 1, or leverage values that were 3 times the average value and,to control for multicollinearity, variance inflation factors had to bebelow 10.

3. Results

The mean duration of pain was 7.2 years (range: 3 months to 40years). Forty two patients and 15 asymptomatic participantscompleted the questionnaires. Of these, 19 patients (45%) wereclassed as moderately disabled (RDQ: 9e16) and 11 (26%) asseverely disabled (RDQ: 17e24), with classifications based on thework by Stratford et al. (1998). Technical problems with the EMGsystem at the start of the study meant that EMG data were notrecorded for three patients. In addition, one patient was excludedfrom the EMG data due to high noise levels in the signal. The EMGdata of two further patients were excluded from the analysis asthey were materially different (>2 SDs) to the values of all otherpatients and they were therefore considered to be outliers.

3.1. Group characteristics

The characteristics of the patient and control groups (whereapplicable) are given in Table 1. The muscle activity figures pre-sented are the average of the four electrode sites used. As shown inTable 1, the control group had significantly reduced muscle tension(as a % of the RVC) (p¼ 0.01) and a trend for greater staturerecovery (p¼ 0.06) compared to the patient group. Variability instature measurements is often used as an exclusion criterion.Analysis was therefore also carried out excluding the staturerecovery data for those participants who found it difficult tomaintain a consistent posture in the stadiometer, as demonstratedby a high SD over the five familiarisation readings. For this purposea SD of 1.7 mm was taken as the cut-off point as it marked a cleargap between the majority of participants and the relatively smallnumber with large SD values, and values exceeding this figure wereconsidered to represent unacceptably high variation. This resultedin the data for 11 patients and one asymptomatic participant beingexcluded. The remaining patients had an average SD of 1.0 mm overthe five readings, which equalled that for the control group. Withthese participants excluded, the control group had significantlygreater stature recovery than the patient group (p¼ 0.03). It wasnoted that the control group had significantly lower body massthan the patient group (p¼ 0.04), which is known to affect bothstature change (Rodacki et al., 2005) and EMG readings (Burden,2008). When the control group was compared with a matched(age, sex, height, weight) subset of the patient group, there wasa trend for reduced stature recovery (p¼ 0.06) and elevated muscleactivity (p¼ 0.09) in the patient group compared to the controlgroup, but it no longer reached significance. This was also truewhen the stature recovery data for those with a SD of over 1.7 mmwas excluded.

3.2. Correlational analysis

The correlations between all markers (excluding EMG data forthree patients as discussed above) are shown in Table 2.

3.3. Correlational analysis: muscle activity

Significant correlations were found betweenmuscle activity andboth pain (p< 0.01) (Fig. 4) and disability (p< 0.01) (Fig. 5). Of thepsychological factors considered, muscle activity demonstratedlinks with functional self-efficacy (p< 0.01) (Fig. 6), depression

(p¼ 0.03), anxiety (p¼ 0.03), pain-related anxiety (p¼ 0.05) andcatastrophising (p¼ 0.04).

3.4. Correlational analysis: stature recovery

Stature recovery was not significantly related to any of the otherfactors. In particular, the relationship between muscle activity andstature recovery was not found to be significant (p¼ 0.17).However, when the 11 patients who demonstrated high variabilityin staturemeasurements were excluded, therewere trends for linksbetween stature recovery and pain (r¼�0.23, p¼ 0.09) anddepression (r¼ 0.25, p¼ 0.08).

3.5. Muscle activity as a mediator in the relationship betweenpsychological factors and clinical outcome

Mediational analysis was performed to investigate whethermuscle activity acted as a partial mediator between any of thepsychological factors and either pain or disability. The results(shown in Table 3) demonstrated that muscle activity was a signif-icant mediator in the relationship between self-efficacy and pain(z¼�1.92, p¼ 0.028). Similar analysis showed that there wasa trend for muscle activity to also be a partial mediator in therelationship between anxiety (p¼ 0.056), depression (p¼ 0.051),pain-related anxiety (p¼ 0.061) and catastrophising (p¼ 0.061)and pain. Muscle activity was not found to be a mediator betweenany of the psychological factors and disability.

R2 = 0.21p < 0.01

0

20

40

60

80

0% 20% 40% 60% 80% 100%

EMG (% of RVC)

Self-

effic

acy

(CP

SS-P

F)

Fig. 6. Relationship between self-efficacy (CPSS-PF) and muscle activity (as a % of theRVC). CPSS-PF: functional subscale of the Chronic Pain Self-Efficacy Scale; RVC: refer-ence voluntary contraction.

Table 4Pain as a mediator in the relationship between muscle activity and disability.

Dependent variable Independent variable R2 Significance

Disability (analysis 1) Muscle activity 0.18 0.007Pain (analysis 2) Muscle activity 0.23 0.001Disability (analysis 3) Pain 0.32 0.011

Muscle activity 0.198

S. Lewis et al. / Manual Therapy 17 (2012) 27e3332

3.6. Pain as a mediator in the relationship between muscle activityand disability

Mediational analysis (shown in Table 4) demonstrated that painwas a significant mediator in the relationship between muscleactivity and disability (z¼ 2.13, p¼ 0.017).

4. Discussion

In line with previous research, there was a trend for patientswith LBP to have higher muscle activity and delayed staturerecovery compared to asymptomatic individuals, although this wasnot significant when comparing to a matched control group, andthe effect size of 0.42 for the comparison of muscle activity (0.71 forthe comparison with the total, unmatched, patient group) was lessthan the average effect size of 1.14 during standing reported ina recent meta-analysis of 20 studies (Geisser et al., 2005). Thepatient group also scored significantly higher on anxiety anddepression than the asymptomatic individuals.

The results confirm that patients with greater pain and disabilityexhibit elevated paraspinal muscle activity compared to those withlower levels. Muscle activity was significantly correlated with self-efficacy, depression, anxiety, pain-related anxiety and catastroph-ising and was found to be a partial mediator in the relationshipbetween self-efficacy and pain. This is an important finding whichverifies the link between psychological and biomechanical factorsin CLBP and appears to confirm the role of muscle activity asa pathway by which psychological factors may affect clinicaloutcome. Although it is widely accepted that psychological factorssuch as self-efficacy have an impact on clinical outcome and there isalso a limited body of research which has reported correlationsbetween such factors and muscle activity, this is one of the firststudies to show that muscle activity acts as a partial mediator inthis way. The role of muscle activity as a partial mediator betweenback pain and self-efficacy (with a trend for a similar role in the linkwith depression, anxiety, pain-related anxiety and catastrophising),suggests that interventions that are able to reduce muscle tension

Table 3Muscle activity as a mediator in the relationship between self-efficacy and pain.

Dependent variable Independent variable R2 Significance

Pain (analysis 1) Self-efficacy 0.24 0.001Muscle activity

(analysis 2)Self-efficacy 0.21 0.005

Pain (analysis 3) Muscle activity 0.29 0.019Self-efficacy 0.169

may be of particular benefit to patients demonstrating thesecharacteristics, which may help in the targeting of treatment forLBP. For example, the presence of elevated muscle tension mightindicate that a patient should be screened for the presence ofpsychological factors as a priority and conversely, high scores oncertain psychological questionnaires might act as triggers to indi-cate that a patient is likely to demonstrate elevated muscle tension.

Although muscle activity was significantly associated witha range of psychological factors, it was not found to be related tofear of movement, which appears contrary to current literaturerelating to the fear-avoidance model and muscle guarding. Thismay be because muscle activity was only measured during relaxedstanding whereas hyperactivity due to muscle guarding, forexample, may become more apparent in certain postures or duringmovements perceived as threatening/harmful.

The results of the mediational analysis suggest that muscleactivity affects disability via its influence onpain, further confirmingthe importance of muscle activity in LBP. However, the relationshipbetween these three variables is likely to be more complicated thana single pathway and, within their impact on pain, both muscleactivity and disability may separately play a mediating role.

The data did not support the hypothesized relationshipsbetween stature recovery and the other factors considered,including muscle activity. This is in contrast to the findings ofHealey et al. (2005a), who did establish such a relationship inindividuals with mild disability (RDQ 5.6� 2.9). The current studyhad the advantage of deriving data from a clinical sample withmoderate levels of pain and disability and the results suggest thatthe relationship between muscle activity and stature recoverywithin this patient population may be more complex than origi-nally thought. However, the results may also reflect the heteroge-neity typical of such a clinical population and the existence of sub-groups within the patient group. In addition, the stature changemeasurements may have been influenced by patients attending atdifferent times of day, whereas Healey et al. (2005a) restricted thetesting sessions to approximately one hour after rising.

5. Conclusions

Patients who demonstrated higher paraspinal muscle activitywere thosewithmore severe CLBP and themediational analysis alsoindicated that muscle activity may affect disability via its influenceon pain. The results therefore support the clinical relevance of thismeasure and suggest that treatments that reduce muscle activitymay improve outcome. In addition,muscle activitywas significantlycorrelated with a number of psychological factors and was found toact as a partial mediator between self-efficacy and pain, confirmingthe link between psychological and biomechanical factors in CLBP.Furthermore, it suggests that there may be particular benefit inreducing muscle activity in those with low self-efficacy.

Acknowledgements

The authors wish to thank both the patients and the staff at thephysiotherapy department at North Manchester General Hospitalwhere the data collection took place.

S. Lewis et al. / Manual Therapy 17 (2012) 27e33 33

References

Adams MA, Bogduk N, Burton K, Dolan P. The biomechanics of back pain. Edin-burgh: Churchill Livingstone; 2002.

Ambroz C, Scott A, Ambroz A, Talbott EO. Chronic low back pain assessment usingsurface electromyography. Journal of Occupational and Environmental Medi-cine 2000;42:660e9.

Anderson KO, Dowds BN, Pelletz RE, Edwards WT, Peeters-Asdourian C. Develop-ment and initial validation of a scale to measure self-efficacy beliefs in patientswith chronic pain. Pain 1995;63:77e84.

Baron RM, Kenny DA. The moderatoremediator distinction in social psychologicalresearch: conceptual, strategic, and statistical considerations. Journal ofPersonality and Social Psychology 1986;51:1173e82.

Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the hospital anxiety anddepression scale: an updated literature review. Journal of PsychosomaticResearch 2002;52:69e77.

Burden A. Surface electromyography. In: Payton C, Bartlett R, editors. Biomechanicalevaluation of movement in sport and exercise: the British Association of sportand exercise sciences guide. Abingdon: Routledge; 2008. p. 77e102.

Coons MJ, Hadjistavropoulos HD, Asmundson GJG. Factor structure and psycho-metric properties of the pain anxiety symptoms scale-20 in a communityphysiotherapy clinic sample. European Journal of Pain 2004;8:511e6.

Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures for chronic painclinical trials: IMMPACT recommendations. Pain 2005;113(1e2):9e19.

Foster NE, Thomas E, Bishop A, Dunn KM, Main CJ. Distinctiveness of psychologicalobstacles to recovery in low back pain patients in primary care. Pain 2010;148:398e406.

French DJ, France CR, Vigneau F, French JA, Evans RT. Fear of movement/(re)injury inchronic pain: a psychometric assessment of the original English version of theTampa scale for kinesiophobia (TSK). Pain 2007;127:42e51.

Geisser ME, Ranavaya M, Haig AJ, et al. A meta-analytic review of surface electro-myography among persons with low back pain and normal, healthy controls.The Journal of Pain 2005;6:711e26.

Healey EL, Fowler NE, Burden AM, McEwan IM. Raised paraspinal muscle activityreduces rate of stature recovery after loaded exercise in individuals withchronic low back pain. Archives of Physical Medicine and Rehabilitation 2005a;86:710e5.

Healey EL, Fowler NE, Burden AM, McEwan IM. Repeatability of stature measure-ments in individuals with and without chronic low back pain. Ergonomics2005b;48(11e14):1613e22.

Jensen MP. The validity and reliability of pain measures for use in clinical trials inadults. Review paper written for the initiative on methods, measurement, andpain assessment in clinical trials (IMMPACT) meeting, April 12e13, 2003.IMMPACT-II; 2003.

Johnston M, Pollard B, Hennessey P. Construct validation of the hospital anxiety anddepression scale with clinical populations. Journal of Psychosomatic Research2000;48:579e84.

Kori SH, Miller RP, Todd DD. Kinisophobia: a new view of chronic pain behaviour.Pain Management 1990;3:35e43.

Lewis SE, Fowler NE. Changes in intervertebral disk dimensions after a loading taskand the relationship with stature change measurements. Archives of PhysicalMedicine and Rehabilitation 2009;90:1795e9.

McCracken LM, Zayfert C, Gross RT. The pain anxiety symptoms scale: developmentand validation of a scale to measure fear of pain. Pain 1992;50:67e73.

Osman A, Barrios FX, Gutierrez PM, Kopper BA, Merrifield T, Grittmann L. The paincatastrophizing scale: further psychometric evaluation with adult samples.Journal of Behavioral Medicine 2000;23(4):351e65.

Osman A, Barrios FX, Kopper BA, Hauptmann W, Jones J, O’Neill E. Factor structure,reliability, and validity of the pain catastrophizing scale. Journal of BehavioralMedicine 1997;20(6):589e605.

Pallant JF, Bailey CM. Assessment of the structure of the hospital anxiety anddepression scale in musculoskeletal patients. Health and Quality of LifeOutcomes 2005;3(82).

Rodacki ALF, Fowler NE, Provensi CLG, Rodacki CLN, Dezan VH. Body mass asa factor in stature change. Clinical Biomechanics 2005;20:799e805.

Roelofs J, McCracken L, Peters ML, Crombez G, van Breukelen G, Vlaeyen JWS.Psychometric evaluation of the pain anxiety symptoms scale (PASS) in chronicpain patients. Journal of Behavioral Medicine 2004;27(2):167e83.

Roland M, Fairbank J. The RolandeMorris disability questionnaire and the Oswestrydisability questionnaire. Spine 2000;25(24):3115e24.

Roland M, Morris R. A study of the natural history of low back pain: part 1.Development of a reliable and sensitive measure of disability in low-back pain.Spine 1983;8:141e4.

Stratford PW, Binkley JM, Riddle DL, Guyatt GH. Sensitivity to change of theRolandeMorris back pain questionnaire: Part 1. Physical Therapy 1998;78(11):1186e96.

Sullivan MJL, Bishop SR, Pivik J. The pain catastrophizing scale: development andvalidation. Psychological Assessment 1995;7:524e32.

Turner JA, Fulton-Kehoe D, Franklin G, Wickizer TM, Wu R. Comparison of theRolandeMorris disability questionnaire and generic health status measures:a population-based study of workers’ compensation back injury claimants.Spine 2003;28(10):1061e7.

Vlaeyen JWS, Linton SJ. Fear-avoidance and its consequences in chronic musculo-skeletal pain: a state of the art. Pain 2000;85:317e32.

Vlaeyen JWS, Seelen HAM, Peters ML, et al. Fear of movement/(re)injury andmuscular reactivity in chronic low back pain patients: an experimental inves-tigation. Pain 1999;82:297e304.

Woby SR, Roach NK, Urmston M, Watson PJ. Psychometric properties of the TSK-11:a shortened version of the Tampa Scale for Kinesiophobia. Pain 2005;117:137e44.

Woby SR, Roach NK, Urmston M, Watson PJ. The relation between cognitive factorsand levels of pain and disability in chronic low back pain patients presenting forphysiotherapy. European Journal of Pain 2007;11:869e77.

Zigmond AS, Snaith RP. The Hospital Anxiety and Depression Scale. Acta Psy-chiatrica Scandinavica 1983;67:361e70.